Hello, I have a small email problem which is a bit difficult to locate, because it appears and disapears again without changing anything on the server. I have a Suse 10.2 / Ispconfig 2.28 Server running and I am hosting about 150 domains on it. Sometimes (very rarely!) my clients get a message like 'max number of connections exceeded' when trying to send/receive mails with an email client. The problem then will persist for about 30 minutes and then vanish without a trace ;-) The last time that happened I made a netstat -tap, wich you can view in the following: Code: Aktive Internetverbindungen (Server und stehende Verbindungen) Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name tcp 0 0 *:mysql *:* LISTEN 2091/mysqld tcp 0 0 *:sunrpc *:* LISTEN 2396/portmap tcp 0 0 *:smtps *:* LISTEN 24900/master tcp 0 0 *:hosts2-ns *:* LISTEN 2913/ispconfig_http tcp 0 0 *:ftp *:* LISTEN 2856/vsftpd tcp 0 0 *:smtp *:* LISTEN 24900/master tcp 0 0 localhost:826 *:* LISTEN 3194/famd tcp 0 0 localhost:48972 localhost:826 TIME_WAIT - tcp 0 0 localhost:36388 localhost:sunrpc TIME_WAIT - tcp 0 0 *:imaps *:* LISTEN 2478/couriertcpd tcp 0 0 *:pop3s *:* LISTEN 2558/couriertcpd tcp 0 0 *:22 *:* LISTEN 2874/sshd tcp 0 0 *:pop3 *:* LISTEN 2550/couriertcpd tcp 0 0 *:imap *:* LISTEN 2479/couriertcpd tcp 0 0 *:www-http *:* LISTEN 3130/httpd2-prefork tcp 0 0 *:smtps *:* LISTEN 24900/master tcp 0 0 *:smtp *:* LISTEN 24900/master tcp 0 0 *:https *:* LISTEN 3130/httpd2-prefork tcp 0 0 whitespace.scr:www-http pd9eb825b.dip0:veracity TIME_WAIT - tcp 0 0 whitespace.scribbl:imap p5089176e.dip0.t-i:vpad VERBUNDEN 4234/imapd [COLOR="SeaGreen"]tcp 0 0 whitespace.scribbl:imap xxx.xxx.189.230%538:58674 VERBUNDEN 25274/imapd[/COLOR] tcp 0 0 whitespace.scribbl:imap p5089176e:simple-push-s VERBUNDEN 22141/imapd tcp 0 3076 whitespace.scribbl:22 p57A4BCDF.dip0.t-:62818 VERBUNDEN 24664/sshd: wartung [COLOR="SeaGreen"]tcp 0 0 whitespace.scribbl:imap XXX.XXX.189.230%538:57825 VERBUNDEN 21789/imapd tcp 0 0 whitespace.scribbl:imap XXX.XXX.189.230%538:57824 VERBUNDEN 21785/imapd tcp 0 37 whitespace.scribbl:imap XXX.XXX.189.230%538:58779 FIN_WAIT1 - tcp 0 0 whitespace.scribbl:imap XXX.XXX.189.230%538:58767 VERBUNDEN 22246/imapd[/COLOR] tcp 0 0 whitespace.scribbl:pop3 p5089ffbf.dip.t-d:64045 TIME_WAIT - tcp 0 0 whitespace.scribbl:pop3 p57bd6320.dip.t-d:49168 TIME_WAIT - tcp 0 0 whitespace.scribbl:pop3 p57bd6320.dip.t-d:49169 TIME_WAIT - The green marked lines are the connections to the client in question. There do not seem to be entries in the mail logs concerning the problem. Any ideas what's going wrong here? Do I have to increase the 'max number of connections' somewhere? Thank you, regards, Tom
Are you really sure that the problem appears when sending mail? The processes you marked above are from your IMAP daemon and not from postfix, so if these processes cause the error message, the error must occur when the client connects with imap. Maybe you can set the limit of simultanious connections to a higher number in your imap client?
Hello. No, I'm not sure. It seems to be rather a problem with opening the imap folders while starting the email client software. I've changed the MAXPERIP value from originally 4 to 30 in the /etc/courier/imapd file to resolve the problem. This seems to fix the problem as far as I can tell now, but I wonder if this might be a security issue. regards, Tom
